The present invention relates to a data processing technique for processing primary data such as input information so as to generate secondary data.
In an inventory control system in a franchise chain or the like, information such as the number of articles sold is inputted from a terminal at each store. The primary data such as the sales number inputted is collected to a data center or the like and totaled. The collected primary data is converted into secondary data which can easily be used by a user for article inventory control.
When each store is connected to the data center online and the primary data is transmitted in real time, there is a danger to increase the network traffic. Moreover, there is a case that realization of online itself is desirable.
To cope with this, such a method can be considered that the primary data is temporarily accumulated and batch processing is performed periodically to generate secondary data. However, such a batch processing has a problem that there is a time difference between the timing of input of the primary data and the timing of generation of the secondary data. It is preferable to prevent generation of a difference between the stock amount registered by the system and the actual stock amount which may cause lack of articles or an excessive stock.
In order to solve these problems, there is considered a method for generating data in accordance with necessity when referencing the secondary data. However, when it is necessary to use an enormous amount of primary data for generating secondary data, the processing requires a plenty of time. Moreover, even when only a part of the primary data is updated, all the programs necessary for generation of the secondary data should be executed.
For this, JP-A-2001-331354 discloses a technique for deriving a relationship between the primary data, the secondary data, and a program required for generation of the secondary data.